Journal of Applied Physiology
Vol. 81, No. 4, pp. 1594-1597, October 1996
METABOLISM

Effect of heat stress on glucose kinetics during exercise

Mark Hargreaves, Damien Angus, Kirsten Howlett, Nelly Marmy Conus, and Mark Febbraio

Department of Physiology, The University of Melbourne, Parkville 3052; and Department of Human Movement Science, Royal Melbourne Institute of Technology, Bundoora, Victoria 3083, Australia

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Hargreaves, Mark, Damien Angus, Kirsten Howlett, Nelly Marmy Conus, and Mark Febbraio. Effect of heat stress on glucose kinetics during exercise. J. Appl. Physiol. 81(4): 1594-1597, 1996.To identify the mechanism underlying the exaggerated hyperglycemia during exercise in the heat, six trained men were studied during 40 min of cycling exercise at a workload requiring 65% peak pulmonary oxygen uptake (O_{2 peak}) on two occasions at least 1 wk apart. On one occasion, the ambient temperature was 20°C [control (Con)], whereas on the other, it was 40°C [high temperature (HT)]. Rates of glucose appearance and disappearance were measured by using a primed continuous infusion of [6,6-²H]glucose. No differences in oxygen uptake during exercise were observed between trials. After 40 min of exercise, heart rate, rectal temperature, respiratory exchange ratio, and plasma lactate were all higher in HT compared with Con (P < 0.05). Plasma glucose levels were similar at rest (Con, 4.54 ± 0.19 mmol/l; HT, 4.81 ± 0.19 mmol/l) but increased to a greater extent during exercise in HT (6.96 ± 0.16) compared with Con (5.45 ± 0.18; P < 0.05). This was the result of a higher glucose rate of appearance in HT during the last 30 min of exercise. In contrast, the glucose rate of disappearance and metabolic clearance rate were not different at any time point during exercise. Plasma catecholamines were higher after 10 and 40 min of exercise in HT compared with Con (P < 0.05), whereas plasma glucagon, cortisol, and growth hormone were higher in HT after 40 min. These results indicate that the hyperglycemia observed during exercise in the heat is caused by an increase in liver glucose output without any change in whole body glucose utilization.

hyperglycemia; catecholamines; hyperthermia

INTRODUCTION

IN ADDITION to the marked circulatory and thermoregulatory alterations that occur during exercise in the heat, there is an increased rate of muscle glycogen utilization (3, 4) that is associated with higher muscle and blood lactate levels (3, 4, 18). These responses are due, in part, to hyperthermia and higher plasma epinephrine levels (3, 10), although circulatory alterations may also play a role. An exaggerated increase in blood glucose has also been observed during exercise in the heat (3, 4, 17) that must be due to liver glucose output exceeding muscle glucose uptake. Indeed, it has been demonstrated that splanchnic glucose output is increased during exercise and heat stress (15). However, it also possible that muscle glucose uptake may be reduced secondary to an accelerated muscle glycogenolysis (5) and/or reduced muscle blood flow. In a previous study in which leg glucose uptake was measured during walking exercise in the heat (13), it was not possible to determine the effect of heat stress because exercise in the heat always followed that in the cool environment. Thus the present study was undertaken to examine the effect of heat stress on glucose kinetics during exercise.

METHODS

Subjects. Six endurance-trained men [20.2 ± 0.4 (SE) yr, 72.1 ± 1.5 kg] took part in this study after being informed of all risks and stresses and giving their written consent. The study was approved by the Human Research Ethics Committees of The University of Melbourne and Victoria University of Technology. Peak pulmonary oxygen uptake (O_{2 peak}) was measured during incremental cycling to fatigue at 20-22°C and averaged 4.63 ± 0.14 l/min.

Experimental procedures. The subjects were provided with food [~14 MJ, 80% carbohydrate (CHO)] for the 24 h before a trial and were asked to abstain from exercise, alcohol, and caffeine for this period. In addition, they were instructed to consume 5 ml of tap water/kg body weight on waking to ensure adequate hydration status. The subjects reported to the laboratory in the morning after an overnight fast on two occasions at least 1 wk apart. On arrival, they voided and were weighed nude, and a rectal thermistor (Monatherm, Mallinckrodt Medical) was positioned 10-15 cm beyond the anal sphincter. Catheters were inserted into an antecubital vein of each arm for blood sampling and tracer infusion. A primed (3.3 mmol) continuous (44.7 ± 3.2 µmol/min) infusion of [6,6-²H]glucose (Cambridge Isotope Laboratories, Cambridge, MA) was then commenced and was maintained during 2 h of rest and 40 min of exercise. The exercise was conducted on a cycle ergometer (Lode, Groningen, The Netherlands) at a workload requiring 65 ± 2% O_{2 peak} in an environmental chamber maintained at either 20 [control (Con)] or 40°C [high temperature (HT)]. Relative humidity was <50% in both trials. No fluid was provided during either trial. Three subjects completed Con first, followed by HT, with the order being reversed in the other three subjects. Subjects sat in a chair at 20-25°C for the 2-h rest period before entering the chamber and commencing exercise at the designated environmental temperature. Venous blood samples were obtained 20 and 10 min before exercise, immediately before exercise, and at 10-min intervals during exercise for analysis of plasma glucose and [6,6-²H]glucose enrichment. Samples obtained immediately before and during exercise were also analyzed for hemoglobin and hematocrit. Additional samples were obtained immediately before exercise and after 10 and 40 min of exercise and analyzed for plasma lactate and plasma levels of various hormones. For plasma lactate, 250 µl of plasma were deproteinized in 500 µl of 8% perchloric acid and spun, and the supernatant was stored at 20°C. Expired gases were collected into Douglas bags at 10-min intervals during exercise for the measurement of oxygen uptake and respiratory exchange ratio (RER). Heart rate (Accurex, Polar) and rectal temperature were monitored continuously and recorded every 10 min during exercise. On completion of exercise, the subjects exited the chamber, were towel dried, had all catheters and probes removed, and were reweighed nude.

Analytic techniques. Oxygen and carbon dioxide contents of dried expirate were analyzed with Applied Electrochemistry S-3A/II and CD-3A analyzers (Ametek, Pittsburgh, PA), whereas volume was measured with a Parkinson Cowan gas meter. Hemoglobin was measured spectrophotometrically (OSM-2 hemoximeter, Radiometer, Copenhagen, Denmark), and hematocrit was determined by microcentrifugation so that changes in plasma volume from rest could be calculated (2). Plasma glucose was measured with an automated glucose oxidase method (YSI 2300, Yellow Springs Instruments, Yellow Springs, OH), and lactate was determined with an enzymatic spectrophotometric method (11). Plasma insulin (Incstar, Stillwater, MN), cortisol (Orion, Espoo, Finland), and glucagon (1) were measured by radioimmunoassay, whereas plasma growth hormone was measured with an immunoradiometric assay (Bioclone, Marrickville, Australia). Plasma catecholamines were determined with a single-isotope radioenzymatic method (TRK995, Amersham). Plasma [6,6-²H]glucose enrichment was measured as described previously (12). Briefly, 500 µl of plasma were mixed with 500 µl of 0.3 M Ba(OH)₂ and 500 µl of 0.3 M ZnSO₄ and spun. The supernatant was passed down an ion-exchange column, washed with distilled water, and dried. The samples were then resuspended with distilled water, placed in glass vials, dehydrated, and derivatized with the use of pyridine and acetic anhydride. The derivatized glucose level was then measured with a gas chromatograph-mass spectrometer (5890 series 2 gas chromatograph, 5971 mass spectrometer detector, Hewlett-Packard, Avondale, PA). Glucose kinetics at rest and during exercise were calculated with a modified one-pool non-steady-state model (16), assuming a pool fraction of 0.65 and estimating the apparent glucose space as 25% of body weight. The data from the two trials were compared by analysis of variance for repeated measures with significance at the P = 0.05 level. Specific differences were located with the Student-Newman-Keuls post hoc test. Paired comparisons were made with a t-test. All data are reported as means ± SE.

RESULTS

Exercise in the heat resulted in a higher rectal temperature (P < 0.05) and greater loss of body mass after 40 min (Table 1). Heart rate was higher throughout exercise in HT compared with Con (P < 0.05). No differences in oxygen uptake were observed between trials at any time point, but the RER was higher in HT after 30 and 40 min of exercise. Assuming a nonprotein RER, total CHO oxidation was higher during exercise in the heat, averaging 234 ± 12 and 201 ± 8 µmol ^· kg^{1 ·} min¹ in HT and Con, respectively (P < 0.05). Average values for heart rate, oxygen uptake, and RER during exercise are given in Table 1. There were no differences in the plasma volume changes during exercise at any point between trials, and the 40-min values are reported in Table 1. Plasma glucose levels were similar at rest in the two trials (Fig. 1) but were higher during exercise in the heat (P < 0.05; Fig. 1). This was the result of a higher hepatic glucose production during the last 30 min of exercise (P < 0.05; Fig. 1) because the glucose rate of disappearance (R_d) and metabolic clearance rate (MCR) were not different at any point (Fig. 2). On average, the glucose R_d was similar in the two trials (Con, 21.8 ± 2.3 µmol ^· kg^{1 ·} min¹; HT, 24.8 ± 2.3 µmol ^· kg^{1 ·} min¹). If we assume that this glucose was fully oxidized within contracting skeletal muscle, it is possible to derive a minimal estimate of the rate of muscle glycogen oxidation from the total CHO oxidation and glucose R_d (14). This value was higher during exercise in the heat (209 ± 11 vs. 179 ± 7 µmol ^· kg^{1 ·} min¹; P < 0.05). Although there was no difference in plasma lactate levels after 10 min of exercise (Con, 3.2 ± 0.3 mmol/l; HT, 3.8 ± 0.4 mmol/l), they were higher after 40 min of exercise in the heat (6.3 ± 0.9 vs. 2.7 ± 0.3 mmol/l; P < 0.05). Plasma hormone levels at rest were not different between the two trials (Table 2). After 10 min of exercise, plasma epinephrine and norepinephrine were higher in HT (P < 0.05), whereas plasma insulin, glucagon, cortisol, and growth hormone were not different between trials (Table 2). After 40 min of exercise, there was no difference in plasma insulin between Con and HT; however, plasma glucagon, epinephrine, norepinephrine, cortisol, and growth hormone were all higher in HT compared with Con (P < 0.05; Table 2).

Table 1. Physiological responses during 40 min of exercise at 65 ± 2% O2 peak in an environmental chamber at either 20 or 40°C Con HT  Mass, kg  0.73 ± 0.04   1.16 ± 0.05*  Plasma volume, %   6.1 ± 2.0   6.3 ± 1.2  Oxygen uptake, l/min 2.96 ± 0.11  3.09 ± 0.10  RER 0.91 ± 0.01  0.94 ± 0.01* Heart rate, beats/min 154 ± 3  175 ± 4* Rectal temperature, °C   Preexercise 36.6 ± 0.2  36.6 ± 0.1    40 min 38.2 ± 0.2  39.1 ± 0.2* Values are means ± SE; n = 6 subjects. O2 peak, peak pulmonary O2 uptake; Con, control (20°C); HT, high temperature (40°C); , change in; RER, respiratory exchange ratio. * Significantly different from Con, P < 0.05.

Table 2. Plasma hormone levels during 40 min of exercise at 65 ± 2% O2 peak in an environmental chamber at either 20 or 40°C Rest 10 Min 40 Min Insulin, pmol/l   Con 48.3 ± 3.3  39.3 ± 5.4  40.8 ± 5.2    HT 54.5 ± 6.0  38.6 ± 4.9  61.7 ± 17.9  Glucagon, pg/ml   Con 33 ± 4  30 ± 4  29 ± 2    HT 31 ± 3  30 ± 2  47 ± 5* Epinephrine, nmol/l   Con 0.13 ± 0.02  0.76 ± 0.08  1.06 ± 0.20    HT 0.23 ± 0.06  1.29 ± 0.21* 2.24 ± 0.40* Norepinephrine, nmol/l   Con 1.98 ± 0.22  7.35 ± 0.70  10.36 ± 1.10    HT 2.35 ± 0.33  10.05 ± 1.03* 18.95 ± 2.88* Cortisol, nmol/l   Con 306 ± 39  332 ± 30  387 ± 34    HT 352 ± 35  367 ± 44  532 ± 56* Growth hormone, mIU/l   Con 2 ± 1  5 ± 4  43 ± 10    HT 4 ± 2  17 ± 6  89 ± 12* Values are means ± SE; n = 6 subjects. * Significantly different from Con, P < 0.05.

DISCUSSION

The major finding of the present study is that the exaggerated increase in blood glucose levels during exercise in the heat is due to a greater increase in liver glucose output without any alteration in whole body glucose utilization. Furthermore, the observation that total CHO oxidation was higher during exercise in the heat, in the absence of any significant alteration in glucose R_d, suggests that muscle glycogenolysis is enhanced under such conditions. This is consistent with the higher plasma lactate levels at the end of exercise and with previous studies (3, 4) in which muscle glycogen utilization, determined directly from muscle biopsy samples, was shown to be higher during exercise in the heat.

Liver glucose output during exercise in humans is regulated by a complex interplay of neurohumoral factors (for a review, see Ref. 6). Although epinephrine may not be essential for the exercise-induced increase in liver glucose output (7), an increase in plasma epinephrine during exercise results in enhanced liver glucose output (7, 9) and higher blood glucose levels (9). In the present study, plasma epinephrine levels were higher after 10 and 40 min of exercise (Table 2), which may account, in part, for the increased liver glucose output we have observed (Fig. 1). Furthermore, the higher plasma norepinephrine levels during exercise in the heat (Table 2) reflect the enhanced sympathetic activity under these conditions, and although direct sympathetic neural innervation may not be essential for the increase in liver glucose output during exercise (8), increased sympathetic activity may also have contributed to the greater liver glucose outptut. Of note, both plasma epinephrine and norepinephrine were elevated after 10 min of exercise (Table 1), suggesting that they may be responsible for the higher liver glucose output during the early stages of exercise in HT (Fig. 1). Plasma insulin levels were not different between trials and therefore appear not to account for the differences in liver glucose output. Although glucagon, cortisol, and growth hormone may not play a major role in stimulating liver glucose output during exercise of the intensity and duration used in the present study (6) and were not different between trials after 10 min of exercise, their increased plasma levels after 40 min in HT (Table 2) could also have contributed, in part, to a greater liver glucose output. Finally, it has been suggested that splanchnic hypoxia can cause increased glucose ouptut from the liver during exercise in the heat (15). We have no data in the present study to either support or refute such a mechanism.

It has been suggested previously that an increase in muscle glycogenolyis as a result of local epinephrine infusion into an exercising limb is associated with a reduction in muscle glucose uptake (5). This effect is believed to be mediated by an increase in intramuscular glucose 6-phosphate concentration ([G-6-P]) as a consequence of the enhanced glycogenolysis (5). Although there may not be a large increase in [G-6-P] during exercise under cool conditions in the present study, an accelerated glycogenolysis during exercise and heat stress (3) may result in a higher [G-6-P] during HT. In addition, if muscle blood flow is reduced during exercise in the heat as a consequence of increased skin blood flow, this could also reduce muscle glucose uptake, although there is some debate on the effect of heat stress on muscle blood flow during exercise (13). A reduction in muscle glucose uptake may result in a lower whole body glucose R_d, thereby contributing to the hyperglycemia during exercise in the heat. However, in the present study, there was no difference in glucose R_d during exercise between the two conditions (Fig. 2), and, if anything, glucose R_d was slightly higher during HT, most likely as a consequence of the hyperglycemia (Fig. 1) because the MCR was almost identical in HT and Con (Fig. 2). Thus we were unable to observe any inhibitory effect of increased epinephrine and/or reduced muscle blood flow on muscle glucose uptake as measured by the glucose R_d. Alternatively, although the majority of tracer-determined whole body glucose uptake during exercise under cool conditions is accounted for by uptake into contracting skeletal muscle (9), this may not be the case under conditions of hyperglycemia and hyperthermia. It is possible that the hyperglycemia observed in the present study (Fig. 1) resulted in enhanced glucose uptake by other tissues. In addition, an increase in skin blood flow and sweat gland activity during exercise in the heat may have increased the glucose uptake in this tissue, although we know of no data on the quantitative importance of skin glucose metabolism during exercise. Thus direct measurement of glucose uptake by contracting skeletal muscle may be required to assess the effects of heat stress on muscle glucose uptake during exercise. Nevertheless, regardless of any potential alterations in tissue-specific glucose uptake, the results of the present study clearly demonstrate that the hyperglycemia observed during exercise in the heat is the result of an increased liver glucose output rather than an alteration in whole body glucose utilization.

In summary, the results of the present study indicate that the exaggerated increase in blood glucose observed during exercise in the heat is caused by a greater increase in liver glucose output without any change in whole body glucose utilization.

ACKNOWLEDGEMENTS

The authors thank Assoc. Prof. Terry Seedsman, Department of Physical Education and Recreation, Victoria University of Technology, Footscray, Australia, for access to the environmental chamber. They also acknowledge the assistance of Vince Murone, Department of Chemistry and Biology, Victoria University of Technology, in measuring the glucose enrichments and of Max Bennett, Michael Christopher, Suzanne Fabris, and Cecilia Peterson with the hormone analyses.

FOOTNOTES

Address for correspondence: M. Hargreaves, Dept. of Physiology, The Univ. of Melbourne, Parkville, Victoria 3052, Australia.

Received 26 December 1995; accepted in final form 5 June 1996.

REFERENCES